#dedicated-servers
1 messages ยท Page 43 of 1
EU law only says you need to protect your network. it doesn't say anything about indirect binding or userspace network address translations to localhost
your multihome is set to 0.0.0.0 so you enabled unreal's internal packet router for when you are bound to more than one address or NIC
๐คฆโโ๏ธ
just an idiosyncracy of the unreal network implementation. it offers no additional security
it doesnt matter to what you set you multihome, it alwas says 0.0.0.0
false. see my logfile
let me prove it
my logfile already does. its -multihome is set to a single IPv4
yes, provided it's set before -log -unattended
almost every existing software DO NOT reroute to localhost Fortunately
it is a backdoor network access security issue
yeah really. the fact that it reroutes packets to localhost is scary since it just punches a hole through your firewall
if it doesnt have very secure filtering on top of that behavior, it could literally get sued for creating a hole in any machine running it
and UE is not developped by heavy security network devs, because it is not its job, so it should stay at its place : binding on what it wants and keep it like that
so, stuffs running there
clearly 10.0.0.24
thats also the ip of it
welcome 0.0.0.0
because it always goes for it
and then 127.0.0.1
it never binds to the external ip
and im used to that behavior from literally every software
you are making a stretch here
unreal uses both. if not, the port wouldnt work
but it indeed uses 8888, because 7777, the defaulkt is blocked
but it works
lot of software do bind to 0.0.0.0 then filter request when told to only respond for some address
lot of software DO NOT REROUTE TO LOCALHOST
and again, when talking about UE5 multi-instance, it not binding literally to the address passed as argument is a BUG
So in my dedicated server, if I've lost hard drives in my inventory due to crashes, I'm basically screwed and can't get all the alternate recipes right?
you can recover from crate or killing your old you
if you cant find it you may use fiscit calculator interactive map to find it
and even restore it
no, also in u7 i was able to eternally remove hardrives
/home/sparxy/SatisfactoryDedicatedServer/FactoryServer.sh -ServerQueryPort=15778 -BeaconPort=15001 -Port=7778 -multihome=127.0.0.1 -log -unattended | grep socket -i
[2023.06.23-12.56.42:744][ 0]LogNet: Created socket for bind address: 127.0.0.1:7778
and without equals sign:
/home/sparxy/SatisfactoryDedicatedServer/FactoryServer.sh -ServerQueryPort=15778 -BeaconPort=15001 -Port=7778 -multihome 127.0.0.1 -log -unattended | grep socket -i
[2023.06.23-12.56.53:704][ 0]LogNet: Created socket for bind address: 0.0.0.0:7778
thank you, please use equals sign and try again
by simple spamming escape in the mam after i checked the options. it closed and never opened the screen again.
got none of the recipes and the drive was gone
Fortunately Thank you
โค๏ธ
I was getting scared
ngl
so it does take the port but not the ip you say?
The drive is in your inventory you have to use it again and wait the 10 minutes
I'm not the developer. I am merely posting my observations
also standard convention of shell scripts that a space equals a =
you see my command and the resulting output
I do that trick to get different options.
nope, any script behave as they want to
it clearly does not process multihome correctly with a space
that is epic's fault, not mine
but you run a shell script with options, they should behave after shell scripts convention, nothing you should be able to change
also it still uses the port normally, that should be especially impossible to fuck it up only in specific cases
unrealengine is a native binary. it's not a shell parser. it parses options however the fuck it wants
it literally gets argv array space separated
or whatever C does in its main function
impossible, because -Port 8888 works and it uses 8888
thats what i mean
it cant be that it only works some time
you're being pedantic. they messed up their commandline switch parsing
are you really gonna make me sign up for the unrealengine source code so I can tell you on which line the multihome parsing is fucked?
but you dont write a syntax for every option manually in the parser?
I have empirically proven that it does not correctly parse multihome
so all options should behave the same
again, I do not work for epic, command line switch parsing is not my responsibility
I merely showed you that it functions as if you didn't provide multihome when you provide it with a space
the how or the why are not really in scope here since we were talking about networking
but the wiki even says with a space :why:
wiki page was made on ue4
and in most cases here it'll still solve the ipv6 problem with a space because defaulting to 0.0.0.0 also works
ah, the new wiki uses = now
It's nonstandard arg parsing since it doesn't take double dashes for words
just use getopt ๐ฉ
unreal also targets windows so I don't think it cares much about POSIX XD
More's the pity
it uses (unsigned int argc,char **argv), so an argument count and a list of string
it passes the space seperated arguments as standalone strings and reference them into argv list
so basically what I said with a count included to determine array length ๐
yep, I was only confirming with how it is directly
so btw you even better know it
(like the detail : it doesnt even understand spaces between arguments)
I mean I don't code native apps but I am a software developer so ๐ don't feel bad โค๏ธ
what do you mean feeling bad ?
oh just re: 'so btw you even better know it'
I just like to give it how it is, so people can better grasp the way the may have skipped or something, so there is no loss
even if agreeing
so yeah. did I mention networking is hard? XD
depends what
it's mostly because people lack generalization examples
and it's a lot to visualize
and documentations are often too evasive
for beginners at least
like "local" "internal"/"external" that are mostly context depend and dont mean the same thing depending on what we are focusing on (in the same context)
but there are also lot of shotcut in the academdy part of it, which does not help at all
some fixed things apparently, then become flexible suddenly, and people get lost
when it was dynamic from the start
so a vocabulary issue perhaps
or meanings changing over time as technologies develop
the network as it is as been made off die and retry and ducktaping
it's not really working generally speaking
so people get closer to something fundamentally functional, but with sometimes convoluted and/or indirect ways, until something less more robust is standardized for this
so lot of thing dynamic was made fixed in the past, and then it would not work anymore as things develops
but academy still show it as "fixed but changing often"
and it often fails
I believe lot of people genuinly dont get networking only because it is very badly understood by most of people teaching it
for instance
I'll give you a fundamental hint : a communication connection, is not a globally defined thing
network address IS local and through gateways by NATURE
trying to forbid people to do NATing and the like is only arrogance
I.E : communication parties don't have to and naturally wouldnt know about the other part local final address
well yes, every node on the network decides how it handles packets however it wants
unless they need to be specifically targeted (link in forwarding for ipv4)
as far as basic Internet Protocol is aware, the link terminates at the edge NAT device. that doesn't mean that NAT is inherently bad, it's just harder to set up for the owner of the NAT device and keep everything working downstream
so I'll drop it here : in a natural network, all parties dont have to get a globally understood address, unless they need to get contacted by anyone on the network)
i.e dynamic addressing can take place as a party ask for connection to a globally understood address
a bit like for anycast addresses/multicast address
like it does with arp table in switch
arp table is to connect layer 2 and 3 of the OSI model. once you go to layer 2 you're no longer talking internet protocol
ethernet frames and hw addresses
I'm mostly spitting onto making a higher layer protocol that ["needs"] to be aware of the ip address
it can even be a security issue as it is not reliable
if it stay into ip : then no problem
layer 2 osi is already a network, I'm happy to see latest scientific papers start to rely less on osi layer model
it is already a network yes, but it's not necessarily an internet protocol network
ofcourse
to be even more transparent : osi layers are pratically a bit shifted
that's why arp is considered layer 2.5
as I said, things that were taught as fixed, are not that clear in the end
yes, because with dynamic routing/switching your switches need to know where the next hop is for a certain IP destination, which is usually cached in ARP. layer glue if you will ๐
Idk if you know, but btw the same mechanism for ipv6 as been standardized as part of ip layer directly, to make it a bit more clear
and so more robust
ye, neighbor discovery
exactly
(sorry it's always a bit complicated to know where people are on the learning curve)
sorry, I have no professional experience in networking so I just know bits and bobs I happen to catch here and there XD
you are very fine, you did your research good ๐
tbh lost of experienced professional have just solidified their knowledge only by working and learning as they would need to further get how it works
I'm too curious for that shit XD I always wanna know how everything works
that is a required skill in technical software/IT ๐
there is always a gap between pratical researchs going on and what is taught in school, and IT is very victim to that
(it really depends on schools and school structures in states etc..)
it is just the beginning of informatic techs
which can be scary or amazing, or both
I would say both
I literally only went to college for the piece of paper that says I know how to do my job XD
I don't think I learned much useful there
:/
.. I aborted college because of that
๐ฃ
I want to get things right and understand, if I must drop that for a paper, then no thank
in one of my courses one of the profs got replaced halfway through so when I went for my exam there was this moment of me and prof not knowing eachother XD
that is to say, I stopped visiting most courses several weeks in XD
I did have good scores, even when skipping courses
I dropped when I was met with a engagement paper (that I didnt sign) and I knew I werent gonna handle the 3,5 next years like that
in the end once you have decent job history they don't care about the paper anymore anyway
it's just a bit harder to get history when you have no paper to start
especially in covid etc
discord ???????
what was that
side note : sorry for everyone coming here, feel free to ask about satisfactory dedicated servers
damned
we dont intend to hijack this channel
that's is too big discord, wtfh, how can one moderate this
I didnt know you could do that, I only put '#'
markdown?
# h1
## h2
### h3
text?
this looks a lot more different on mobile
Has anyone run into the latest experimental dedicated server just hanging when you connect?
I am working on a friends server at the moment and its not apparent what is going on, but I just copied his kubernetes deployment/service etc to my home and it works just fine
``` is what we are getting when trying to connect
is NAT. are the ports forwarded correctly?
that is port, server query port and beacon port UDP
I'm like 99% sure its not networking related
its for both internal and external users
ports are forwarded, and we can see the user connect in the server logs
do you have a new save or an existing save?
I also checked outbound and reflection to make sure my buddies firewall isnt the issue
brand new
we nuked the volume and started from fresh to rule out
you can try disabling game pausing from the server manager
I'll double check, I know its not paused on my end, but buddies I didn't watch
I did ask about that though
dedicated server pauses by default when nobody is connected
it's just a setting you can change when logged in as admin
ya, i remember its part of the image we are running as well as on setup
I'll double check with him once he is off his work call here in a few
so you can connect to the server manager fine, it fails when you try to join the game? U7 or U8?
its U8
and yeah, it goes to the loading screen with the top of the day etc on it
it sits for a bit
then goes back to main menu with a timeout error
HMMMMMM
try setting -multihome=pod ip or -multihome=0.0.0.0 in the launch parameters, before -log -unattended if you have those
I might have rubber ducked myself, but I could be wrong
multihome is set to 0.0.0.0
it shows the client network is the overlay net though, so I am wondering if buddies CNI is jacked up
nope, it shows on my end too nvm
it's 'cause it's NAT
can you share the deployment config?
so I can't grab buddies config, but I'll send mine
its the same, just environment changes
buddy running k8s in the cloud/dedi server or some local homelab setup?
buddy can connect from LAN but not from WAN?
buddy can't connect from LAN either?
deployment in my lab works all the way around
his lab doesnt work at all
internal or external
have him try spin up an nginx pod or something first, see if connectivity works at all maybe ๐
he is on the same vlan/sub as his lab too so not a lot of nat other than cni
it does, we have Ark servers running there too
its REALLY weird
the only thing that is MAJORLY different is I run BGP
like, bot my ELB setup where as he is simple L2
GOT IT
it was pause simulation
"hey man did you turn pause off? yes. are you sure? yes. are you sure that you are sure? YES!"
I just had him check on stream, it was on xD
Thanks for the help
no problem, seen more timeout issues with pause being turned on. networking issues would usually be server manager not loading either, unless there's a typo in a port config somewhere. and k8s networking config issues are outside my ballpark XD
right lol
I tried talking to some others as well, and all I ever heard was like check the NAT, its def NAT
I mean to be fair, for most home users trying to run satisfactory server, NAT is often a good guess XD
true
I am like tell me that to my dedicated HA firewalls, and my 4 node kubernetes cluster, and the L3 networking stack running FRR
don't think the usual homegamer is running that xD
not really no ๐ I'd go broke on electric cost alone
don't remind me
once you mentioned k8s I was considering it had to either be a bug (i.e. it was, game pausing being borked), or some esoteric issue I wouldn't be able to help with anyway other than rubber ducking
ok you can host ๐
Does anyone have the issue where trees come back but cant be destroyed now? Is there a workaround by any chance?
is there commands to use to change play cap and tickrate? I can't find this game.ini file that everyone is talking about
are you on windows or linux ?
windows epic games :)
(for the server)
installation folder + FactoryGame\Saved\Config\WindowsServer\
30 tickrate is the default
what do you mean game folder ?
what is the name of the file to change the player cap?
someone mentioned the game.ini file I guess that's something else then
it's Game.ini
you need to select show option : file extension
to better naviguate in configuration files
was thinking that
otherwise it can be misleading
you need to add :
MaxPlayers=6```
is this good?
cool, tysm man
it may be safer to make the file readonly after that
so the server doesnt override it when it restart
makes sense, I'll do that
also it is better to edit these file when the server is completely offline
yeah it is dw
then you good
yes multiple people have this issues, apparently you can do a manual save then restart the dedicated server, and it's better after that, but I'm not sure, it would be better to restart client games as well ofcourse
can you use a a secure tunnelling service like ngrok for a dedicated satisfactory server
Hello, i want join server satisfactory, but i have two errors :
Invalid authentication token
Your connection to the host has been lost
hello i have problem in my save, my game always crash in 1 minute playing another people have solution
it's possible
can you help me pls
i have 76 hours of playing
Thanks il try that
probably since the server supports forwarding
did you try to load it on your game directly (as local game) ?
sometimes you can load it locally, then save it locally then upload it again and it fix some problems
also did you make all the common checks like having enough free RAM etc ?
yeah i trying to load in local game
and in 1 minute my game crashing always
i dont understand
so it crashs both when in dedicated server and when in local game ?
yes exactly I launch or join the game and my game crashes between 30 seconds and 1 minutes after launching
... has anyone ever received an in-game notice saying the server will restart in 30 minutes? o-o
some do apparently, some people said it was when an update is avaiable
an update for what specifically though
looks like the session restarted without doing anything else in particular
performance and rubber banding still crazy bad, any recommendations yet ?
So running satisfactory when i connect to it on the query port it says that is offline and not authenticated? Is there a way i can fix this?
where do I find the save for the dedicated server (my old pc), also has anyone had issues with the save corrupting or anything because I'm planning on doing a daily back up
on linux it is in the .epic folder, so i assume on windows it would be in %localappdata%\.epic so so, i dunno google it
got the same problem .. ๐ฆ
howto fix this
Hey everyone, tldr server on Mac. it's been a hot minute since I've been in here (Update 5ish..) and I saw some mentions about the server running on Mac M1 (arm) via Win11 VM. I'm planning to try it, but if you've had success and know any tricks I'd appreciate feedback ๐
Hello! Can you please tell me if there are any problems with installing Dedicated server in Experemental now? I want to try running it on another computer? Is it worth it now or is there a big problem with it. I run the server for 2 people to play, when playing through my computer there are some bugs, for example, the vehicle constantly breaks down and I have to disassemble and reassemble it. Are there even more problems when playing multiple people on a Dedicated server or +- the same way?
I have not seen that since update6. I believe it happens when the server does basically a manual save and upload. I think was on a default interval like every 24 hours or 48 hours.
doesnt happen for us at all. so its at least not default.
i guess some third party stuffs defaults or so
i have never seen it in hundreds of hours on dedicated
first time I've ever seen it. don't have anything special configured. uptime was around 24h tho
in U7 we never made 24h uptime due to dismantle crashes ๐
linux ~/.config/Epic/FactoryGame/Saved/SaveGames/server
windows %userprofile%\AppData\Local\FactoryGame\Saved\SaveGames\server (I guess, maybe a bit different at the end)
Hey guys - should I be able to connect to a dedicated server running on the same box as my client using the loopback address? I'm trying but having no luck, and not really sure why
hm - after playing around, it works if I specify -multihome=0.0.0.0 but does not work if I leave off the bind option. I'm guessing something about the dedicated server doesn't like my IPv6 config?
the dedicated server in its current state dont like lot of stuff and behavior weirdly why network binding options (-DisablePacketRouting and -multihome)
we are a lot to try to get when it works and when it doesnt
tldr : currently it is like the client were trying a new address when the connection is already established on another address, if you are on local machine you could just try ipv6 loopback though ::1
ah interesting, yeah if I do -DisablePacketRouting that works too
(which makes sense since I see that -multihome triggers that behavior)
it's a bit blurry
I as well as some other people think it is a mismatch into the network manager dedicated server addon and the new UE5 behavior (different from UE4 used in update 7)
we just are sad it is so complicated to connect the server since it may prevent experimenting the other things
yeah makes sense
anyone else noticing the server kicking you when you dismantle large blueprints?
Hi! I got a dedicated server from Shockbyte and the lag is terrible. I'm trying to redownload the save file from the server so I can just continue to host a game on my computer for my friends to join. I cannot find the save file from the FTP server. Does anyone know how I would go about getting the save file that has been worked on while on the dedicated server?
its a linux server ?
^
~/ is the satisfactory user home
I don't have that. I only have servermanager.sav. Might this be it?
On my PC I do not have a file named "server" under \SaveGames. On the FTP for the server I do not have any save files, and it only shows .log and .cache in \SaveGames.
what is the root of your ftp
on your pc you are not supposed to find it as the server doesnt run on your pc
but in the ftp it must be there
Let me try downloading Filezilla rather than doing anything from their web-based FTP.
I dont know if it will give you any access
because ftp is not really secure, and web based can just be locally driven on their side
I meant : in what folder do you start
Does someone knows a lot about setup a dediscated server.. It's been 3days i try to make it work but it's not working;.
Explaination :
I got a server runnin in Satisfactory Update 7 Early Access. I got to reinstall my entire server because of a bug on it. After reinstalling Ubuntu Server, nothing worked anymore. I verified the firewall (ufw), the installation of steam and the game, the dependances, i tried all the thing to make Satisfactory Early Acces or Experimental work, but without success.
What i get :
On Early access server : Instant crash, Segmentation fault on starting.
On Experimental Branch : It starts, but in game, i got TIMED OUT "20.01" and i can't find anythink to change the value of the server timeout
Have a wonderful day/night,
Zarcross.
(ping me for reply, thanks โค๏ธ )
This is as far back as I can go in the FTP
I just saved too, and none of the "last modified" times are recent (other than the world.zip backup I just created)
ok so the .conf @subtle sedge
then Epic
then FactoryGames, etc
is that what you went for ?
Ahh yup I'm stupid. I went directly to FactoryGame in the home directory. Sorry! I got it. Thanks for your help!
probably the network bug everyone are facing (in u8) also if you run it at home, make sure your port forwarding rules still applies to the machine
you can try adding -DisablePacketRouting -multihome=yourserverlocalipaddress
(in the server starting script arguments at the start)
OMG thanks you a lot ! Experimental U8 works now
Hmm on public network it doesn't seems to works.. Local OK but timed out on my public IP
but DisablePacketRouting makes it work on local network
I tried both, like, i tried u8, then u7, then again u8 and again and again trying differents configurations
But yea, it's possible that i installed u7 first
you said your server has a local IP and a public one. did you port forward correctly in your NAT device?
all three of the ports, udp proto
oh you mean that i have to forward 15000 and 7777 with the "15777", i forwarded only 15777 and i worked before, but i can try
(it is on UDP ๐ )
you need to forward the server query port (server status), the beacon port (connection handshakes) and the server port (game). so 7777, 15000 and 15777
ok i'll try that and i tell you if it works ;D
uh, do i have to put my local ip or my public ip on the "-multihome="
ok, i think i made a mistake somewhere on my NAT device and delete the two others records, now it works on both ip, tanks you a lot for your help, you are the best :D
so it works well with that setting now ?
You're very welcome. Good luck on Massage-2(A-B)b, pioneer!
(so we know as feedback)
what is that thing
what is what? XD
"Massage-2(A-B)b"
The name of the planet you play on
so it works well on both public and private ip with -DisablePacketRouting -multihome=YourServerLocalIpAddress now ?
fr, never noticed
it's in the lore ?
yes
damned
I think during landing it says
"good luck on your mission on Massage-2(A-B)b pioneer" something like that ?
Yep on that config it works ! I think it was my mistake + the "-DisablePacketRouting" that are the problems
Thanks you a lot !
but you still have it on right ? (I want to make sure that setting can work with public ip too)
yep i put the two flags on my sh script
so u8 server with multihome settings would be network fixed now Yay
๐
I feel better
thank you
๐
Thanks YOU ! for your help for the community
I hope some others players can be helped by that !

you can see it in the codex too normally, but mine is blank on the server XD
I imagine getting DC'ed while in a truck is not an uncommon experience?
opening this one instantly crashes the server

Anyone got an idea if there's an nssm command/arguments that shows the console?
Can yo even call console in the main menu?
hey guys, how can i change autosave interval?
Server is crashing every autosave ...
Got autosaves every few minutes...
one is possible with command, one only in config files.
not sure if its interval or amount.
i think interval was possible ingame
mayve try with a =
nop, i also dont find anything about the config file ...
okay,, just looked it up in the config file there it is changed to 1800 - but game still does autosave every 3 minutes
did you estart the server?
lol, many times
where are the savegames on linux dedicated servers?
ah lol
full location is ~/.config/Epic/FactoryGame/Saved/SaveGames/server
got it yeah, but is experimental that hard bugged in dedicated servers ?
We get only Time Outs, Rubberbanding, etc.
On Linux, Windows .. Doesnt care.
it also depends on your setup and server location, so if you are running localhost on your windows machine it will work better network wise, and if your server is a bit slower in the cpu you can get more rubber banding
I recommend using -DisablePacketRouting -multihome=YourServerLocalIpAddress
(in the server starting script arguments at the start)
so networking is fixed
im hosting the server .. never had issues, except for that experimental branch atm.
I disabled package routing already.
with the multihome , stupid question lol. Do they mean private or public ip there ? ^^
but yeah cpu is a bit slow with 2,1 GHz lol
the one on your server machine, as I wrote in the line
what does disablepacketrouting exactly do for the game?
so private ok, but with multihome set, packagerouting should get disabled automatically i think?
console command is broken in U8. set in config file manually while server is stopped. #dedicated-servers message
yeah, did that ๐ But is it seconds or minutes there ? ^^
When i changed that it always got reset to 30 after restarting
set it read only
did that then ๐
the while server is stopped bit is important. it flushes its runtime config to disk on exit. expected behavior
is it recommended to change the the network options there ?
This ones ..
some people say it can help.
but need to be changed on server and client right ?
you say the server crashes on save or do all players time out?
Save -> 30 seconds lag -> TimeOut
do you have a large save?
freshly started map ... 900KB size
that's weird. if you look in the logs, how long does the save take?
i also set the internet speed and it removed most rubberbanding etc.
also thats a client problem
because even when the server terminates completely, you can still move freely before timing out
so you dont ever get a lag and then timeout from something happening server side
do you know, for what i should search in the logs?
give me five. I need to get to my computer to check
probably just the word saved
saved in x
i think
this one i also get much
0,4 seconds to save ๐ So shouldnt be the problem there
don't think you're disconnecting due to the saves
saving seems normal
I've had it happen when saves take over twenty seconds that your client drops connection but this is not that
okay
so you need your client log for this now
but what should i look up there
think im going to try this .. Change file, set to read only, restart ?
well, what exactly happens when yoz timeout
yo is there any new servers i can join
Hey so i wanted to play on a dedicated server with friends hosted by one friend but if i try to connect it only displays "This server apperas to be offline"
can the other friends join without issues?
yep
remove the server and add it again, make sure to double check that the IP and port are entered correctly
i have tried that and changing firewall settings, restarting game and pc, reinstalling satisfactory already
I'm not sure. Could indicate an issue with your outbound firewall rules, possibly in your router, if you have one with such advanced functionality
I've never seen straight up connection failures that were client side, only server side config issues. but you said others can connect externally so idk
ill try to connect to another network
via my phone
if i pig the domain it comes back with "Destination net unreachable." strange...
can you tell me the domain? ๐ค
not sure gotta ask the friend
sounds like either the dns is misconfigured, or the domain record was very recently created and still needs to propagate, or the dns record points to a bogon network
you can try connecting on the IP address directly instead of the domain
i just noticed something if i force ping via IPv4 it works because the dont have an IPv6 addres yet so is there any way to force IPv4 in satisfactory
lemme try
you can disable ipv6 in your computer's network adapter
yeah but i need IPv6 for things i do work releated
the only reason it would try to connect to ipv6 is if your friend configured ipv6 on the domain name (i.e. an AAAA-record)
manually enter the ipv4 address instead of the domain name in the server manager
yeah when using the Ipv4 addres it works
tell your friend that if they don't have ipv6 properly configured, delete the AAAA-record for the domain ๐
ill do thanks for the help!
no problem ๐ first time I've seen this issue but it's good to keep in mind for me that this can happen too XD
i think thats something many peopel can have issues with
it's usually more of an issue on the server side since clients usually connect with an ipv4 or a domain name that only has records for ipv4 XD
Why does a satisfactory server lag so much ?
a minecraft server does not lag on my side
If you have any tips
because I have 5ms with the serv and it lags enormously
what's the tick rate if you look in the server manager?
When trying to connect to my server i just setup i'm getting this after a few seconds. Any ideas as to what would help fix this? https://i.imgur.com/JmlByoR.png
can you try launching the server with -multihome=192.168.0.250 ?
Yup 2 seconds
do you get the error on join or does the server manager not load at all?
@clear lagoon Yup that seemed to be the problem not sure why...
someone an idea hat this is?
its quite spammy
it tends to break on windows with ipv6
haven't really looked into why because I personally don't use windows XD
I had just installed the server on another pc locally and was going to connect to it on server manager and then that is when i received that error. not even connecting to join the server
I don't use ipv6 ๐ฆ and im pretty sure i disabled it on my ubuntu server
ip a
and check if it says ipv6
sorry, no clue, that specific entry doesn't appear in my server logs
Hi,
I tried to generate a map on my dedicated server but i have these errors :
do you know what i can do ?
Sorry for my english i'm french ๐
So I checked and it does say ipv6 however i do have it disabled in the settings menu
You can ignore those as long as you do not experience any adverse effects
Actually the problem is that it doesn't generate a map ๐ฆ
I'm not sure what you mean by that. Does it not show in the server manager or can you not join the game?
I create a map there (first photo) :
I set a name, when i want to connect i have this message (second photo) :
and my VPS is crashing at this point
are you running U7 or U8? how much RAM does your VPS have?
I'll try to find the answer for the first question, for the second, I have 4GB, not enough maybe ?
I think it needs 12 - 16, believe that is whats said on wiki
no, for the stable version (update 7) you you need at least 12GB
for update 8 (experimental) you can get away with less (6-8)
Ohhhh okay so that's it I think, thank you ๐
Does anyone know why I would receive this message:
When trying to just setup a new Experimental Branch server?
Out of context, my thought is they take a comparison between two components and square them [(a-b)^2] but in the case of 0.0 a and b are equal, likely duplicates (race condition kinda stuff) - possible packet loss so local variable isn't updated. Or I'm completely wrong lol
Can anyone tell me where the script is on Linux and how to add them exactly?
I wanna run the dedicated server as service but somehow the disable packet routing argument doesn't work in the service conf
"The best way to tweak those values is by running FG.NetworkQuality <number from 0 to 3> from the server console until we expose this in the server settings UI"
Am I dumb or whats going on? :D
What is your service config? Because it worked ok when I made a systemd unit file for my server
I don't see anything obviously wrong, although you do have some things that I don't have. Not sure if one of those is causing it. Here's what I have for my unit file
it does not
Thank you, I'll try that version
it works for me :
..
ExecStart=/srv/Satisfactory/FactoryServer.sh -DisablePack....
..```
(I installed with steamcmd +force_install_dir /srv/FactoryServer)
I figured the path. Weird. Maybe the double quotes are wrong at execstart?
I assume they aren't needed (or else mine wouldn't work lol) but idk if they actually will screw anything up if you quote the ExecStart command
I have never quoted ExecStart (or other commands) when I wrote unit files before, fwiw
ExecStart lines arent shells, they are parsed by systemd, so quoting it entirely is wrong, for instance
Thanks for the input. The sources on the web for creating services aren't that easy sometimes
My guess is that maybe by quoting it, you're causing systemd to look for a binary named "/foo/bar/FactoryServer.sh -DisablePacketRouting"
yep
that would be that, for instance
if you quote the full line
to make things simple :
- you want systemd to separate arguments correctly
- you want the satisfactory shell to not see quotes as arguments
so avoid extra quotes
I did. Still get the connection timeout. D'oh
When I loaded the script manually it worked
in which file is the multihome line? I couldn't find anything
the multihome is an argument
like I showed
so in this case just the argument for the script?
Should Loopback work? otherwise I have a problem
so it looks like :
ExecStart=/srv/Satisfactory/FactoryServer.sh -DisablePacketRouting -multihome=YourServerLocalIpAddress -ServerQueryPort=15777 -BeaconPort=15000 -Port=7777
wait local ip. okay so the one the router gives the machine
sorry didn't see the local
You should be able to connect via the loopback address if that's what you mean. I can connect to my server by going in-game and adding 127.0.0.1
when multihome is not set I think though
and only when you are running the server on your current machine
ah, yeah can't speak to that. I found that DisablePacketRouting was enough for my purposes so I don't have multihome set
you still can connect using the server local ip
its the ip on the server machine
you can use ip ad show to see it (iproute2 network tool package you are expected to have on your linux)
at least that's the name of the package on archlinux
sorry
I forgot the ad I feel stupid
I didn't know you could abbreviate it further, I've always typed ip addr show lol
Fun fact, I just tested and you can even go as far as ip a show
Used ifconfig. Was easier for me
or even ip a s, wow
I'm on a debian derivate
linux is trying to push people to switch to iproute now
but debian and derivative still use the old way
You aren't. I learned more about systemd from you than from the interent. Thank you very much for your help
And thank you very much also
np man
ip l a l eth0 nif t v i 654 the "they fooled us" meme incoming
For what it's worth the best resource I know of for systemd is actually the man pages. Kind of a slog but more accurate than people's random blogs or whatever
(dont use that command please)
and more important : usually up to date
sometimes things are deprecated, and going learning from a 5 year old article can be double edged
(= ip link add link eth0 name nif type vlan id 654)
lol, I was wondering what some of those things expanded to
I got as far as "ip link add link eth0" before I got lazy and said "yeah I can imagine this would be fun"
So the startup command for the service now is:
/home/gnzlbrmpf/SatisfactoryDedicatedServer/Engine/Binaries/Linux/UnrealServer-Linux-Shipping FactoryGame -DisablePacketRouting -multihome=192.xxx.xxx.xxx
Still geht the connection timeout๐ช
you are trying to launch the shell directly
the package include some script that set it up first
Nope.
One sec
[Unit]
Description=Satisfactory dedicated server
After=network.target
[Service]
ExecStart=/home/gnzlbrmpf/SatisfactoryDedicatedServer/FactoryServer.sh -DisablePacketRouting -multihome=192.xxx.xxx.xxx
Type=simple
Restart=always
User=gnzlbrmpf
[Install]
WantedBy=default.target
And that's still not working?
No it isn't
Silly question but did you remember to systemctl daemon-reload after changing the file?
I did that, yes
Also stopped the service and restartet it
server was offline on the client
could a reboot help?
Are you running the client locally or on another box?
client and server are two different machines
And on the client machine I assume you're able to otherwise reach the server - like if you ping 192.x.x.x it works
Do you have a firewall running on the server? If so is it allowing the satisfactory ports?
firewall is inactive
extended version of what I'm using is
#dedicated-servers message (there are all the 3 files above and below it)
with installation as :
(errata: add systemd (linked above) 'satisfactoryd-update.service' file too)
if you want to check some specific things, that is all the details of my installation
can dedicated servers be ran on update 8 yet?
Your server isn't behind a nat or something is it? I am assuming when you ssh you use that same 192.x.x.x IP
yes, it could be run on u8 day one
Thank you!
be aware it is extra experimental though
extra!
tbh I'm going though a dyndns to make sure port forwarding works. The only open ports the servers is listening to are the game ports and my ssh/rdp port
If I start the FactoryServer.sh -DisablePacketRouting manually the server works fine
ok well that at least helps rule out a lot of network config issues
oh and my setup file is currently with
OTHERSETTINGS="-DisablePacketRouting -multihome=ServerMachineLocalIpAddress"
(server-TheNameOfTheInstance.conf file)
dyndns does not setup port forwarding
it only give you the ip to connect to
I guess I'm not sure of why there would be any differences network reachability-wise in a manual start from your ssh session, vs if you start it via systemd
The server should spit out messages to the journal, maybe try journalctl -xe and see if you spot any errors that look promising?
if it is same context (not other container or anything) then it is supposed to be same network yep
The port forwarding is set up in my router. It worked fine wirth update 7
and u8 as well
the 3 ports are forwarded ?
(you can drop TCP, and keep UDP only)
Our server is crashing specifically when we deconstruct items in our base, can be any item, and it is seemly random. But the server crashes about 10 or so times a day if we play during the day full time. Here is a snippet of the crash log if anybody has any ideas on it.
<RuntimeProperties>
<CrashVersion>3</CrashVersion>
<ExecutionGuid>A2B092F563CC4C259BD09A2431236901</ExecutionGuid>
<CrashGUID>UE4CC-Linux-993B71C7A79445EE9C4AD2B688D365B7_0002</CrashGUID>
<IsEnsure>false</IsEnsure>
<IsAssert>true</IsAssert>
<CrashType>Assert</CrashType>
<ErrorMessage>Assertion failed: (Index >= 0) & (Index < ArrayNum) [File:Runtime/Core/Public\Containers/Array.h] [Line: 674] Array index out of bounds: 2550 from an array of size 2550</ErrorMessage>
<CrashReporterMessage />
<ProcessId>38</ProcessId>
<SecondsSinceStart>74563</SecondsSinceStart>
Good to know thanks. Any suggestions on the linux distro to use for the server? Seems like I have to reinstall it cause the FactoryServer.sh quits itself now. And yes, I stopped the service beforehand. Will try a reboot and if that won't work I'll reinstall the whole machine. Right now it's on LinuxMint
I'm running it on Arch cause my desktop is Arch (btw), and it works great on that. But... I wouldn't recommend you use Arch just for the hell of it lol
I tried arch. Some commands are different and stuff but the heck can you customize it
Yeah I gave it a whirl just for the sheer learning experience
It was kind of eye opening to have to build the system up from basically a kernel and nothing else
I would not recommend change distribution just for trying to make a dedicated satisfactory server working on it, unless you are already familiar with it, especially stuff like archlinux which even if it can be great, require a lot of "customization":tailoring
I never was able to run the game itself on any linux (don't know why) so I was somehow stuck with the debian derivates cause they're easier to learn^^
it's like buying tailored cloth, you NEED to go get your mesures and stuff
I don't mind customization tailoring or the learning curve.
My general linux recommendation is basically Ubuntu or a derivative
Cause they seem to be the most "it just works"
(which you're already on)
I'm archlinux team but only if you like the taste of blood or reading lot of documentation
Manjaro just works as well and is arch
taste of blood? Gimme. Documentation -> with enough time always
I'm used to the manpages
unfortunately Manjaro do not really work as well :c
for the standard user they usually do.
man the arch wiki is definitely kind of great
Ok, I'm not the standard user so I guess that's why manjaro didn't work for me lol
plain arch?
the down side is : sometimes it works greater for other distribution than it does for archlinux, because when debian and unbuntu switch to the "new schemes" wikis are already not up to date for archlinux
yes plain arch, manjaro add tools that are usually duplicate of archlinux ones, and sometimes it conflicts, it says it review arch packages before running it on manjaro, but in fact it sometimes just break dependencies of other packages that already went downstream but depends on the new stuff that is not there yet, archlinux rolling when things like that happens can be 3-4 days to settle, but when manjaro in between, manjaro tools do not always understand it and it is half updated for like 2 weeks sometimes, which breaks things for people using it
manjaro is nice, but Imo it should more stick to being a archlinux preset setup and pre added tools that are customized to work with each others, than trying to be isolated from archlinux repositories and tools
Is it safe to update the system with the server running? I know the rolling release but also had some downs with these distros
depends what you are updating, that's why pacman will always ask you if you commit to it or not
that's why archlinux need some knowledge to run smoothly and other thing around that line
the thing manjaro is nice that archlinux doesnt do : isolate kernel packages
that's the major update downside about archlinux : if you update kernel you better be shuting down then restarting
because it overwrite the kernel packages, with the currently executing said kernel older version's part that are still not loaded
if any software want it, it will try to load new modules into the current kernel and it will fail
manjaro have everykernel side by side, it takes more space but you can update safely
I still dont really get why archlinux does not do that, when it could then have cleanup hooks to not stack too many versions
that especially happens with graphic updates
archlinux really is rolling release, it is only made to run on the edge
(but you can set it up to run older package, but you need to do the fancy adapting settings by yourself)
kernel updates can bepainful but most distros keep the old kernel. At least the ones I know. On Mint you have to delete the old ones manually
Which can cause problems as well
the thing is on arch the kernels are ~3 to 9 days lagging behind the literal linux stable branch release
so it is updated very very often
I sometimes run kernels that are literally 2days old
The question for me is: Are they stable enough for SF server? ๐
In your three systemd files:
Do I just need to adjust the paths/username or is there a need to change the variable calls as well? When using your install instructions?
certainly, the tricky part is making a stable configuration that fit your needs at the start
then it will be reliable, unless you update breaking things (like new edge breaking python super update version and you rely on it)
The server is just for SF. I don't need python or anything, just ssh, maybe rdp and kate^^
if you follow the instructions you can use these files as they are exactly 1:1, I put instructions on creating satisfactory user
you can still customize it though
Thanks, I'll give it a shot then. Maybe the server will work afterwards
like changing satisfactory user to satisfactory-game user everywhere in the files and instructions
I'd just change the username
the only thing important to change is your starting username
that's why I put a weird username as placeholder
Everywhere. That's clear. Otherwise it won't run
so you can even sed 's//' instruction.file into it
Don't worry, I understand the files and will read them attentive before using them. But they are a great headstart
you may also want to setup -u and -g numbers of the user and group if you want to use graphic login on the machine
otherwise it will use >1000 uid and may show up on login screen
Let me check if ssh is working on android then the others won't need to access the server as long as my phone has an internet connection
(I prefer using ssh on a machine with a keyboard personally, I use android only as backup)
the best option is I believe to put the graphic users above something like 5000 id, and then set the graphic login to only gather above that
it gives more room and prevent using system predefined ids
It works, so no graphic login needed
anyway : good luck o/ โจ
Thank you
ps : users of manjaro : setup ssh before needing it, you will escape a lot of troubles when graphic will go off, and you will at least be able to gracefully rollback
ps: raw or cook yourself (or raw with your fine custom sauce :D)
I hope I didn't kill this channel
some of these bugs are fixed on u8
unfortunately it is a breaking update, so it's either not further developed (u7) as is or experimental (u8) as is
most bugs like that are fixed in u8, but u8 introduce new bugs
they plan on stabilizing u8 along september if I recall correctly, as a lot of the dev team will be on holiday this 2 months
@minor harbor sorry for ping
errata: you should have 'satisfactoryd-update.service' file too at lines 7 and 38
also here is informations on my server-InstanceName.conf current file #dedicated-servers message
or use this corrected version (also removing notices about things I didnt test that is now tested) :
btw : blood moon :D (or maybe only sightly ?)
Added nicer explanations, giving hints for shutting down instances too
Also here are the needed systemd files (messages link, for completeness) :
#dedicated-servers message
ps: users of arch based distros, use the satisfactory-server AUR package it makes life very easy (you might be using it already, i didnt scroll up just wanted to do this here)
When the game goes live, what happens to the experimental builds?
Thanks. I'll use the corrected verion ๐
it becomes the early access too
And we don't know what will happen after version 1.0. I don't think there are plans beyond that yet since the priority is to release version 1.0.
aur is aur, I improved it a lot for me, you can check the message just above yours to see how
I think the early access branch will just stop being early access on epic and a new final branch will show on steam
I think the same. But I'm not sure if the EA steam version will just be the final one as well. On steam you just have one verion to choose.
like experimental becomes = early access
early access will be = final
until some other update is in development (but the game is considered complete gameplay already)
Still waiting for the announcement of the flushable toilet DLC๐คฃ
Hey Community,
I currently hosting a Update7 Server and i want to check out ser Update8 so i deceided i create another instance and i installed Update8 fresh on the same maschine.
I started it but sadly it does shutdown itself after a couple of seconds while its booting
This is what i can find in the logs
[2023.06.26-07.26.08:274][ 0]LogOnline: Warning: STEAM: Failed to initialize Steam, this could be due to a Steam server and client running on the same machine. Try ru> [2023.06.26-07.26.08:274][ 0]LogOnline: Display: STEAM: OnlineSubsystemSteam::Shutdown() [2023.06.26-07.26.08:274][ 0]LogOnline: Warning: STEAM: Steam API failed to initialize!
and
[2023.06.26-07.26.08:273][ 0]LogGenericPlatformMisc: Error: SetEnvironmentVar not implemented for this platform: SteamAppId = 526870 [2023.06.26-07.26.08:274][ 0]LogSteamShared: Warning: Steam Dedicated Server API failed to initialize.
oh and i forgot to mention , i want to run V7 and V8 simultaniously
and yes i do not use the default ports on V8
i get this when there already is a server running.
so either even with changing ports you cant run two, or thy arent actually changed
Did you change all 3 ports? Have you tried running with -nosteam if you think this is the cause of the failure? Check the CrashReports dir to make sure you're not crashing due to some other issue
yes all 3 ports are changed.
I will try -nosteam now
if that doesn't work, feel free to share the crashreport context xml
okay -nosteam didnt work
and if it's not a crash, share the full log file
[2023.06.26-09.54.17:826][ 0]LogServer: Error: Failed to initialize server. Could not bind any addresses to port '15000'.This is very likely caused by another server instance already running. Use a different port
make sure you set your ports differently than the U7 server. i.e. -ServerQueryPort=15001 with equals signs, not spaces
@signal jasper: ^
so its true i cant run 2 servers even with different ports
then its not using my prompt wich i give with the start of the server via command line
i may try do set the port in the .ini
no no you misunderstand. the other two ports are correctly configured. but the first one is not
you have this set: LogInit: Command Line: -โ ServerQueryPort=15000 -โ BeaconPort=15001 -Port=15002 -nosteam -log -unattended
however ServerQueryPort=15000 won't work as it is the default BeaconPort, so is already in use by the U7 server
set ServerQueryPort to a different number, i.e. 15003
ohhhhhh damn ... there is the mistake !! i mixed the numbers up ^^ ill give it a try
look at that , now it works.
thanks @narrow raft. I really looked over this simple error :p
No problem, glad I could help :)
i have one more question then i am happy ^^
i want to use the Update7 savefile for my Update8 Server but i cant find the savefile on debian.
i cant find it here ~/.config/Epic/FactoryGame/Saved/SaveGames
any idea ?
Hi! Anyone knows how to change AutosaveInterval in last game version? Command from Wiki not working
Also last time I checked it it weren't that good out of the box
there was missing systemd temp/ source directory connections and things
and it also doesn't really supports update
ok, now I'm really thinking it is bad at that point
It should be in ~/.config/Epic/FactoryGame/Saved/SaveGames/server
Specifically, the home folder of the user the server is running as. if you're on linux that is
on windows it's in %LocalAppData%/Epic/FactoryGame/Saved/SaveGames/server if memory serves
damn you are a genius 
i dont know why but today i make really stupid things .. linux user over 10 years but i checked the ~/.config/Epic/FactoryGame/Saved/SaveGames with my admin user and not my user for game instance 
nice!
fair enough, i havent tried updating it yet. However, it worked fine out of the box for me so not sure what your problem with temp was
I dont like it creating a new root directory too
I checked it again so I don't throw spurious point of view :
by default it triggers temp file service on the whole satisfactory directory
I don't like that, it's cleaner to set the right of satisfactory directory directly
and the fact you need steamcmd to create the package then install the package but can't update it without going manual maintenance overwrite mode (you loose benefit of having a package)
doesn't make much sense to me
I would call my aur package "satisfactory-server-servicing" or similar, and put steamcmd as a dependency
I would also put the same logic under steamcmd aur if I was making it tough it self updates so it probably doesnt make much sense
(I plan to open an aur someday btw but I'm waiting a better email setup for managing an arch-Linux account)
and I have a satisfactoryd-update.service
but apparently it's the only way to do it without setting uid ??
then mb on that
that's weird to me
it would be cleaner to use an echo command instead of packaging a literal file though
anyway, it could maybe be nice if I was creating a PKGBUILD for my setup
So was playing on my server and it randomly lost connection and now i respawn as a new character. I can't remove the buggy i was in or retrieve my items any way to fix or clear this?
https://streamable.com/92acgp
"I can't remove the buggy i was in or retrieve my items any way to fix or clear this?" what do you mean ?
also the other character is you in an other session, you have to wait it times out I believe
in worse case you could get the .sav file and edit it with fiscit interactive map
Try doing a manual save and load save. Should put you back into the first body.
hello friendos, we got a server with 4 vcores and 16gig of ram. Everything is good server side but the server is getting giga lag spikes - any suggestions would be appriciated
do we need to gib more cores?
nope 4 is fine, 16g RAM is fine too especially for starting saves, are you sure that's the server that spikes and not the clients ?
did you check autosave settings ?
it usually spikes when saving
Does the server update automatically and restart or is there amyting that the admin needs to do?
I usually close the server and manually update it
might as well, next patch is over a month away anyway
in a way it could automatically update with some fancy setup on the server machine, but that in itself require administration
so if you set it up yourself you can assume it doesnt update automatically
There is a second me i would assume stuck in the buggy and i can't remove the buggy or get in it. I have waited for the character to timeout and it still hasn't been removed. I've also restarted the server.
https://streamable.com/bxr0n5
Like this? or am i doing something wrong? https://streamable.com/5filb3
Yes I don't use vehicles but that is how I get people to spawn back in the correct body's. Vehicles have been pretty buggy.
you are on update 8 ? did you updated your server recently ? some fixes have landed for vehicles
Yeah i had just updated it when i got back after my first message still nothing though so i went in the interactive map and removed the vehicle and just said fuck it to the items i had :/
Just went to check if the body is still there...
you can kill it, to recover the crate
I'll try that in 2 seconds i did smack it a bit but didn't seem to change but ill go beat the shit out of it now
Success!!
Now thats some space savings damn
Well twice i've had this issue now in 1 day... Character lost again have to go farm him after removing the vehicle. Think thats it for me till the server becomes more stable
and for instance that's exactly what is happening rn
the package still isnt officially published by archlinux though, not even on testing repository yet
so currently the dedicated server doesnt need 8G+ RAM anymore?
Anyone getting a weird bug where the game loads and you can look around but you cannot move?
Anyone knows that?
Yes, you need to stop the server and while the server is stopped, change it in config
In GameUserSettings.ini
[/Script/FactoryGame.FGGameUserSettings]
mFloatValues=(("FG.AutosaveInterval", 120.000000))
It's in seconds.
if you have existing mFloatValues, you separate as follows:
mFloatValues=(("some other setting", 123.450000), ("FG.AutosaveInterval", 120.000000))
What command shall I use for it?
Oh, thx a lot!
you can't change it using commands in U8 at the moment
memory use depends on your save size, amount of players, how far they spread out, etc. but you can get away with less RAM than in U7. I think you can still hit around 8-9GB on a large save
did you log out or timeout in a hypertube?
k, because when U6 I was getting > 8g ram usage for a new save so I'm wondering if they optimized the ram usage after then
Vehicles are a little buggy on dedicated server right now for U8, so I suggest avoiding them if you want to use it in U8. If you ever time out or disconnect, wait up to your configured timeout setting (usually 30 seconds by default) before logging back in. you get a new pawn if you log back in before the old one disconnects
it's specifically only gotten better in U8, presumably incidentally due to the new world partitioning system
Good morning everyone
today i have another problem to solve ^^
I just updated Server and Client but i cant join the server. I got the server running yesterday but i couldn had the chance to connect my client because of work and getting home late so i just tried it now with the new update and this comes up
do you have a pretty large save by any chance? just like the client does in singleplayer, the server freezes while performing an autosave, default every 5 minutes. you can check how long saves take in the server logs
is -ServerQueryPort=15010 launch parameter set on your server?
(if you're getting this while trying to connect on the server manager)
yes - because i run 2 servers (U7 and U8)
so i used different ports
i tried other ports too , just checked wich are still free on my root
do you run a firewall on that machine? did you allow port 15010 (and the other two ports) UDP inbound?
yes thats all clear , its an rootserver wich i use for a couple of games . all ports are clear and forwared
is it running on windows or linux?
debian linux
do you have a -multihome= launch parameter set?
no because i saw no use , is it necessery
if not, can you try setting it to -multihome=<server NIC ipv4> ? i.e. from ip -4 a
there are sometimes issues if you don't set it and there's IPv6 addresses present on any of the interfaces
okay ill try ๐ 1 min
you are kidding me - multihome was really needed ...
it works now!
Okay i really would not think that this would be the solution ๐
when I'm bored enough I might at some point go and try to make ipv6 work on purpose XD
but until then I'm not entirely sure why this breaks the server ๐
but if it works it works ๐
๐ ๐
Thanks so Much Lena !
No problem ๐ Good luck on Massage-2(A-B)b, pioneer!
Nope, standing on the ground
Seems to be some kind of latency issue
Even though the ping is 3ms
Fixed itself after restarting the game
you should be able to move once you log in unless your pawn is in a glitched state. have you tried rest-
ah
Seems to be a recurring issue
Sometimes logs me in as a new player
Then I have to go hunt my clone
It's when you log back in too fast, before the old session has timed out. wait for your configured timeout time / look at the server status and wait until the player count drops
Weโve only just started the save 48 hours of game time! We get insane server lag when using hyper tubes as an example. (No one can craft or drag from their inventory while someone is tubing)
However we are running U8 so it could also be that
I don't think I've seen that before. are you using default settings? what's the server tickrate like when someone is using hyper tubes?
hypertubes are also laggy on U7 dedicated servers. But without the crafting problem while someone uses it.
not when small save, my ~160h save hits 7.5GB
16GB RAM still is safe
12GB still is recommended minimum
but you have room until big save
our 135h save takes around 4 gb ram
Hello,
I have started playing Satisfactory again with a friend after a long time and rented a dedicated server from Shockbyte. At first everything was working fine, but when my friend tried logging in today after downloading an update he kept losing connection (I can still join). We have already increased the timeout time but to no avail.
Thanks in advance!
You also have to increase timeout on server side fot it to work
yeah they are, late game it lags so much you just die in the tube
Trying to run as a service on my fresh ubuntu LTS instance and getting this error. Followed every step of the wiki doc to a T. Anyone know what this issue is?
Nevermind, lol, I see part of the issue
your_user
Ok, fixed that line in the unit file for systemd but still getting an error. Its changed a bit though.
Why are you appending +quit? Been a hot minute since I messed with the dedis so I don't remeber
Thats what I meant
In that case might be a plethora of issues
What error does your friend get?
First night testing experimental u8 server. Damn its laggy ^^ 2 Players and the server consol is full of orange warnings ^^
"your connection to the host has been lost"
If anyone knows how to integrate Satisfactory into Discord please let me know where I can find that information. At this point if Satisfactory could post the number of players on a Dedicated server would be awesome! Thanks
so it quits after updating
Ah yeah that makes sense
steamcmd have its own interactive mode otherwise
is there anything i can do so i have better performance on the u8 server ? its really laggy ^^ Server is chilling at 2% -10% cpu and under 10%ram , so its not an performance issue ^^ console is full of warnings ^^
this game is all about single core speed so if you playing on like an old xeon or anything that is not new with high enough IPC then you will experience rubber banding issues etc.. i been running on a 3950X and i dont have any laggy issues.. beside the game being buggy with cars etc but thats something else..
yes i know , but on u7 i dont have any problems on my dedicated server . its a multiplayer problem.
solo also no problem on u8
well then that a particular problem with your setup... but to answer your question no there is nothing you can do to increase performance on the server side beside changing network quality (which i dont think does anything imo) or change tick rate you can try that.. but i haven't change any of those and i dont have any laggy issues and playing with 4 players on the server.. it all depends on your set up
sorry im pretty sure its not my setup.
When i or my buddy laggs ingame i see alot of warnings in the server consol - like audio not found etc (see screenshot above) so i guess there is something wrong with the game serverside.
and again , do you play u7 or u8 ?
u7 is running like butter
i see the same thing all the yellow warn dont mean anything they been there forever now
tickrate is something i can check
im on update 8
and i just shutdown my second server instance running u7 - maybe it helps . lets wait a couple of minutes and see the performance , maybe thats the reason
well if your running two satisfactory server on an old CPU then yeah that would cause the issues...
here are my specs and no issues so far even with two servers running update 8 but i also isolate them with docker so i choose what thread they run on
but the consol really prints out some weird warning ๐
yeah that been a thing since EVER i ran dedicated server since they been release and those warning have always been there
im southercali if you want i can host the save for you on my server and you can test it yourself but if your in the UK like most people are in here then latency will be an issue lol
its our first run since they brought out the dedicated server. So the Satisfactory Ded server is new for me ^^
Ah thats a little far away ๐ Im Central Europe ^^
yeah figure everyone on here is usally from europe lol
Yes cpu is '14 but single core perf. is not much lower.
Its one my best running machines for over 6years (last restart 4 years ago ๐ )
but its weird that u7 dont make any problems but u8 does ... I just updated it today with todays exp. update. Maybe this was the cause ? Did you try it with your friends today with the current update
but , now after 10 minutes we see a little bit less lags. Maybe it was because of the second server.
they did change a lot comapre to update 7 how the world is loaded etc.. but im pretty sure is your cpu IPC is not fast enough to run more than 1 from what i been reading.. you might have the ram to run it but CPU is almost a decade old now
yeah we have i mean the game is still buggy like with the car issues.. and having duplicate characters etc... but is satisfactory server never get any love lol..
dont do cpu shaming ๐ Its my workhourse ^^ Its runns so much stuff but i get your point :p
hmm yea i guess the time will tell
and if your running on a VM you get even more a performance hit i run it docker so i can limit the resource is uses or set it to run on my fastest core on the CPU.. but hey man i dont CPU shame is just facts lol..
it was just a joke ^^ i know its old but it was never a gaming server , my whole company runs on it , and alot of other stuff ^^ and because its running so brave i install some gameservers when i have the motivation to play again ^^
No VM , just pure hardware
but what i just found out ... regardless of the meaning of errors. Everytime one of the players lagg on the server this warning spams around 10times
[2023.06.27-19.00.12:224][966]LogAkAudio: Warning: UAkGameplayStatics::StopActor: Could not retrieve audio device.
only when one of us gets a lag spike for around a second
yeah for gaming server is more about IPC and RAM i always see people on here complaining about rubber banding lag when saving the world etc.. all of that has to do with people playing old xeons with low clocks but high core counts.. or playing on payed server that have the same issue old xeon with low IPC compare to anything new..
yeah there is an issue with the audio deivce i see it too but if the serve is running then ignore them..
that how is ALWAYS been since they got release if you see yellow warning ignore them if the server is running
you can see on here is that game is still very single threaded heavy.. but i would see my whole CPU it would be less than 5 percent.. but as you can see satisfactory still love that single thread speed... so if you want to make the best satisfactory server 7800X3D 64GB+ ram and you have the best gaming server out there for hosting lol...
Ah no i dont care about the best satis server , i would never giveup my workhorse - too much work is in there^^ just was wondering why u8 is so much different then u7 and then i see the warnings wich i dont have at u7 and so on i was wondering whats up ^^
Lol yeah I get that I was trying to point out a rig that would be perfect for just game hosting..
I bet you would see a single core spike up when players are connected if you look at it long enough lol
believe me or not , i have it on all the time because of the laggs currently , there was no single spike over 40% - and we are currently 2 players active playing
for a game hosting that would be really nice , you right
I Believe you but you also have to take into account IPC, Memory bandwidth as your running DRR3 there are so many variables.. but if you tried everything you can do then is just hardware age and satisfactory DS is pretty heavy to run to have a smooth experience from what Iโve see
just wanted to throw in my cent that processor architecture/ipc really does matter, especially for applications that rely heavily on a single thread (i.e. satisfactory simulation loop)
I had it on a Xeon E5-1650 v3 which is positively ancient by now and a 3MB save started clocking in at about 12-13 seconds to complete. at that point I switched to a ryzen 3600 which dropped it the same save to 7s, and now at 8.5M it's still only just under 11s. tick rate is much more stable still
had a stopover on a server running a Xeon E3-1270 v6 but that's not really worth talking about
cpu cache is probably important too
^^^ And thatโs what A LOT OF people donโt understandโฆ
Everything matters IPC/architecture/ cpu cache/ core clocks/ ram speed, bandwidth etc all of that matter in terms of performance
But my main point was that from what Iโve seen on here all the time is people that are complaining about rubber banding, or lag issues, or train or belts rubber banding etc is mainly people who are running the server on old hardware or rented services with old hardware.. and if you did everything you can and still having issues like that then is most likely hardware age
yeah my existing machines weren't really intended for videogame servers so once I started noticing issues I just moved it up to the newest boy I had till I got a new one spun up XD
and a lot of them cheap hosting providers use these ancient second-hand or third-hand dedicated dogpoo that got decommissioned five years ago out of an enterprise environment
๐
the thing I believe is also the game as been optimized for current gen architectures, the engine too and it has been optimized to work with latest helper library I think also, having that much uptime I believe you run older libraries as well, which may not help either
there was a time were server cpu had only 1 to 3 MB of cache, now latest consumer cpus have like 75MB, that's another story for instance
same with ram bandwidth
ram latency have only dropped by nearly half time through 3 DDR generations
but bandwidth have grown to like 8 times the throughput
cpus are monsters compared to old gens
latest ryzen is able to push 8 operations in a single pipeline/thread every cycles
and it goes up 5GHz
All my Systems gets a routine update every month so im not on the oldest liberies , but i get your point.
I just dont have the needs to restart it. Its my root for my small company (homepage, services ,etc) , friends , and some projects. Everything runs fine. I hostet over 20game server of the latest games betweend 2019-2023 and never had any problems like now in u8 satisfactory , so i think its normal when i ask if there is maybe something other the problem then my hardware.
But i get it , its the hardware ๐
of course, I mean after that long running maybe a restart could refresh some stuff, .. or break everything because maybe a lot of your services are not tested against host restarting then
call me old , but with this machine i really live by saying
never touch a running system ๐
most of the services i did has an autostart after crash etc , but i think so too that most likely some of them wont start
but thats where the fun begins
dependency cycles
can still be dodged when crashing
but not when starting
as an example I had a socket streaming service that was giving sockets to dns services, but it was also doing dns requests to connect at the start
it runs fine with cache when some of them crash
but when everything had to start, I had dns failed because no socket and socket failed because no dns
can i ask for help with an issue?
And yeah you are correct most dedicated server like for other game are pretty light.. I just think this game requieres a new cpu to run smoothly.. why because they probably havent optimized themโฆand like they say still DS is Very experimental so thatโs another big reason why is so heavy to run imo
dontasktoask
ask for help directly, otherwise people can't successfully help you, and messaging you messages like I'm doing is still needing additionnal time from us all
ok so where's the issue , my game won't start , at first it said i don't have enough Video Memory to run the game (which is impossible , I am running a 1080 card) i managed to fix it but now every time i try to run a save it just freezes and crashes to desktop with no error
my drivers are up to date and i did verify files
im running out of things to check before re-installing the game
is it a laptop or a desktop ?
Desktop
also are you sure you are talking about video ram ?
because dedicated server doe not need video
That's what it said not enough video memory
looks like you are talking about the game not the server, which is just a bit further up in the channel list #satisfactory-experimental
or you really mean the server ?
i mean the game
if you dont use a dedicated server then #satisfactory-experimental is a better channel to help you
oh, you did go into it first
Lol yeah it seems like no other channel on this whole discord provide support by people they just talk about the game and show thing off LOL only we do I guess LOL
premium support in here; fixed my issues for sure - now I just hang out to see if I can help now and again ๐
oh and I wait for my satisfactory patches to download heh
Hi yall, I put in several hours with a friend but they dont wanna play anymore are there any public servers to tag in and help?
there is the #looking-for-group-old channel with form that takes a line about "- local host multiplayer
- 24/7 dedicated server" choice
thanks ๐
Anyone know if there's any automatable tools that provide players a live semi-always-up-to-date map of the server, similar to satisfactory-calculator.com? I could theoretically do a curl request against a discord webook and dump the most recent autosave into discord and have them individually upload to that site for a map if they want, but wondering if anyone has built out any automatable solutions where they just go to a URL or whatever.
Surely such a tool would only sync every X minutes or whatever, that's fine.
isn't SCIM running entirely client side though?
To my knowledge for that website, you still have to go upload the save file every time
I'm talking about this site here: https://satisfactory-calculator.com/en/interactive-map
Where you can see not only your builds, but resource nodes and other stuff
yes. it's running entirely client-side still. source code is over at https://github.com/AnthorNet/SC-InteractiveMap
you're just loading the save into the web app
Ah, cute. Thanks! I'll poke around with this
yes, just keep in mind that people loading the url will just be downloading the save from your server and loading it into SCIM on their local computer. depending on the bandwidth of your server, how fast the user's computer is and how large the save is (SCIM gets slow on large saves ๐ ) it might not be the best experience
That's not a huge deal, the server is way overprovisioned for this use case lol
it won't be anything like minecraft's dynmap
good luck ^^ do let us know what the experience is like, I'm curious myself now ๐
I will! I'm a sysadmin IRL but this is not quite in my wheelhouse technically, but adjacent enough to skills I do use all the time, so I should be able to push through and get it working.
anyone have tips for cutting down all the client crashes while on servers?
driving vehicles onto the ramps kicks me out of the server every time?
is it a known issue or just for me?
I crashed in an explorer and getting back into the game and getting my stuff has been taking me an hour.
thats true
I believe you could run the code into a nodjs app server-side directly
here you are inviting people to modify their saves and re-upload it, which is not very fair
I've been struggling to launch a dedicated server on experimental using my Steam Deck, I cannot for the life of me figure it out, even after straight up purging the system of anything got to do with a satisfactory dedicated server. It keeps saying it's finding deprecated files but I've scanned my entire system and the folders it complains about don't even exist.
It worked previously before experimental, though I only used it for a short time then
there are a lot of warnings etc.. that are not the problems
there are always these warnings, because it's a cut down version of the game, so lot of the files effectively are missing but not needed
lemme boot my SD back up and I'll have a look through everything that seems relevant
how much free ram do you have ?
the reliable way is to check htop
god I wish I knew what that meant
Also, I can't tell if it's crashing or if something else is going on, because the system monitor says that UnrealServer-Linux-Shipping is using 465.5MiB memory and the game client server browser doesn't say the server is offline
However, it gets permanently stuck saying Server name pending, and then after a while gives me this popup
yep but u8 dedicated server have currently a network bug
(copypast)
I recommend using -DisablePacketRouting -multihome=YourServerLocalIpAddress
(in the server starting script arguments at the start)
as long as you have the 3 ports opened in firewall it should be fine
make sure to get the server local ip though
you can try with 0.0.0.0 instead (bind any)
but I'm not sure it works as great